HPC Guru (Twitter)
RT @thoefler: Prof. David Keyes from @KAUST_ECRC concludes his great @siag_sc talk on approximate matrix representations with a nice tiled-precision block format in #HPC. This is very related to quantization in #AI/#DeepLearning. We should learn from each other :-)!
https://siag-sc.org/supercomputing-spotlights-webinar-series-3.html https://twitter.com/thoefler/status/1587820473945145348/photo/1
RT @thoefler: Prof. David Keyes from @KAUST_ECRC concludes his great @siag_sc talk on approximate matrix representations with a nice tiled-precision block format in #HPC. This is very related to quantization in #AI/#DeepLearning. We should learn from each other :-)!
https://siag-sc.org/supercomputing-spotlights-webinar-series-3.html https://twitter.com/thoefler/status/1587820473945145348/photo/1
insideHPC.com (Twitter)
Check out Silicon Mechanics' article: Overcoming Challenges to Deep Learning Infrastructure
https://wp.me/p3RLHQ-nx8
@ExpertIncluded #HPC #HPCAI #AI #deeplearning
Check out Silicon Mechanics' article: Overcoming Challenges to Deep Learning Infrastructure
https://wp.me/p3RLHQ-nx8
@ExpertIncluded #HPC #HPCAI #AI #deeplearning
High-Performance Computing News Analysis | insideHPC
Overcoming Challenges to Deep Learning Infrastructure
Designing an infrastructure for DL creates a unique set of challenges when the training and inference steps of DL feature distinct requirements.
HPC Guru (Twitter)
Recommended viewing:
Trends in #DeepLearning Hardware
by Bill Dally, SVP of Research @nvidia
EECS Colloquium @UCBerkeley
https://youtu.be/4F_vfPFNe04?t=404
#AI #HPC #GPU
Recommended viewing:
Trends in #DeepLearning Hardware
by Bill Dally, SVP of Research @nvidia
EECS Colloquium @UCBerkeley
https://youtu.be/4F_vfPFNe04?t=404
#AI #HPC #GPU
YouTube
Bill Dally - Trends in Deep Learning Hardware
EECS Colloquium
Wednesday, November 30, 2022
306 Soda Hall (HP Auditorium)
4-5p
Caption available upon request
Wednesday, November 30, 2022
306 Soda Hall (HP Auditorium)
4-5p
Caption available upon request
HPC Guru (Twitter)
RT @thoefler: 2.5 bits per weight are enough to run inference on extremely large language models! Using a simple blocked scaling scheme and moderate block sizes.
Meet GPTQ: https://arxiv.org/abs/2210.17323
#DeepLearning #TinyML @DAlistarh @elias_frantar @AshkboosSaleh https://twitter.com/thoefler/status/1599653546483408896/photo/1
RT @thoefler: 2.5 bits per weight are enough to run inference on extremely large language models! Using a simple blocked scaling scheme and moderate block sizes.
Meet GPTQ: https://arxiv.org/abs/2210.17323
#DeepLearning #TinyML @DAlistarh @elias_frantar @AshkboosSaleh https://twitter.com/thoefler/status/1599653546483408896/photo/1
HPC Guru (Twitter)
In a fireside chat at #GTC23, @NVIDIA’s founder and CEO and @OpenAI co-founder Ilya Sutskever (@ilyasut) discussed GPT-4, #ChatGPT, deep learning’s future and how it all began
https://t.co/8B1BHyYHPp
#AI #DeepLearning #GPU
In a fireside chat at #GTC23, @NVIDIA’s founder and CEO and @OpenAI co-founder Ilya Sutskever (@ilyasut) discussed GPT-4, #ChatGPT, deep learning’s future and how it all began
https://t.co/8B1BHyYHPp
#AI #DeepLearning #GPU
NVIDIA Blog
AI Opener: OpenAI’s Sutskever in Conversation With Jensen Huang
In a fireside chat at GTC, NVIDIA’s founder and CEO and OpenAI co-founder Ilya Sutskever discussed GPT-4, ChatGPT, deep learning’s future and how it all began.
insideHPC.com (Twitter)
Quantum on Wall Street: JPMorgan Chase and QC Ware Report on 'Deep Hedging' with Quantum-Classical
https://t.co/hD2UCjfYV7
@jpmorgan @QCWare #QuantumComputing #quantumclassical #financialservices #HPC #supercomputing #DeepLearning
Quantum on Wall Street: JPMorgan Chase and QC Ware Report on 'Deep Hedging' with Quantum-Classical
https://t.co/hD2UCjfYV7
@jpmorgan @QCWare #QuantumComputing #quantumclassical #financialservices #HPC #supercomputing #DeepLearning
High-Performance Computing News Analysis | insideHPC
Quantum on Wall Street: JPMorgan Chase and QC Ware Report on ‘Deep Hedging’ with Quantum-Classical
New York and Palo Alto, CA — March 30, 2023 — JPMorgan Chase and quantum and classical computing software and services company QC Ware [...]
HPC Guru (Twitter)
RT @thoefler: Submitting a #DataScience, #AI, or #DeepLearning paper to @Supercomputing? Check the performance fallacies in "benchmarking data science".
Twelve ways to lie with statistics and performance - just to make sure your work doesn't fall for any of them ;-).
http://htor.inf.ethz.ch/publications/index.php?pub=473 https://twitter.com/thoefler/status/1643120712474038272/photo/1
RT @thoefler: Submitting a #DataScience, #AI, or #DeepLearning paper to @Supercomputing? Check the performance fallacies in "benchmarking data science".
Twelve ways to lie with statistics and performance - just to make sure your work doesn't fall for any of them ;-).
http://htor.inf.ethz.ch/publications/index.php?pub=473 https://twitter.com/thoefler/status/1643120712474038272/photo/1
HPC Guru (Twitter)
ICYMI: Designing network support for High-Performance #DeepLearning Systems
Network and memory bandwidth is the new oil in #AI #supercomputing
Keynote by @thoefler at #HPC Advisory Council, April 2023, Lugano, Switzerland
https://www.hpcadvisorycouncil.com/events/2023/swiss-conference/pdf/day2/ETH.pdf
ICYMI: Designing network support for High-Performance #DeepLearning Systems
Network and memory bandwidth is the new oil in #AI #supercomputing
Keynote by @thoefler at #HPC Advisory Council, April 2023, Lugano, Switzerland
https://www.hpcadvisorycouncil.com/events/2023/swiss-conference/pdf/day2/ETH.pdf
HPC Guru (Twitter)
#DeepLearning pioneer Geoffrey Hinton quits Google
@geoffreyhinton says he has new fears about the technology he helped usher in and wants to speak openly about them, and that a part of him now regrets his life’s work
https://t.co/MGw895E3wX
#AI via @techreview
#DeepLearning pioneer Geoffrey Hinton quits Google
@geoffreyhinton says he has new fears about the technology he helped usher in and wants to speak openly about them, and that a part of him now regrets his life’s work
https://t.co/MGw895E3wX
#AI via @techreview
MIT Technology Review
Deep learning pioneer Geoffrey Hinton quits Google
Hinton will be speaking at EmTech Digital on Wednesday.
HPC Guru (Twitter)
RT @EuroCC_project: 📢 Workshop Alert! 📢
The 3rd @FF4EuroHPC #workshop on #HPC, #ml, and #DeepLearning in #industry is this Friday!
Come hear about 5 success stories spanning sectors from #manufacturing to #medicine.
🗓️ 9 June, 9:00-12:00
👉 https://www.ff4eurohpc.eu/en/news-and-events/news/2023053016061698/ff4eurohpc_oc2_success_stories_premiere__workshop_3
RT @EuroCC_project: 📢 Workshop Alert! 📢
The 3rd @FF4EuroHPC #workshop on #HPC, #ml, and #DeepLearning in #industry is this Friday!
Come hear about 5 success stories spanning sectors from #manufacturing to #medicine.
🗓️ 9 June, 9:00-12:00
👉 https://www.ff4eurohpc.eu/en/news-and-events/news/2023053016061698/ff4eurohpc_oc2_success_stories_premiere__workshop_3
www.ff4eurohpc.eu
FF4EuroHPC OC2 Success Stories Premiere - Workshop #3
FF4EuroHPC is a European initiative that helps facilitate access to supercomputers and all high-performance computing (HPC)-related technologies for SMEs
HPC Guru (Twitter)
Velocity Bench is an open-source benchmark suite that covers various industry domains including #HPC & #DeepLearning
The goal is to provide a fair performance comparison across #GPU platforms and parallel programming languages
https://t.co/LhR0j9F2WP
#HPC #AI via @HPCwire
Velocity Bench is an open-source benchmark suite that covers various industry domains including #HPC & #DeepLearning
The goal is to provide a fair performance comparison across #GPU platforms and parallel programming languages
https://t.co/LhR0j9F2WP
#HPC #AI via @HPCwire
HPCwire
The New Velocity Bench for GPU Offload Performance Data
Performance benchmarking is the hallmark of HPC. One need to look no further than the Top500 list which has been recording HPC performance since 1993. Historically the Top500 list is […]